Northwestern University supports the concept of a randomized clinical trial to determine risks and accuracy of chorionic villus sampling. The sole unit in the Chicago area in which investigators are both geneticists and obstetricians, and the largest prenatal diagnostic center in the region, we have a well-defined referral base that will facilitate identification of women early in pregnancy who are willing to be randomized. Since March 1984 we have been monitoring continuing pregnancies, and by January 1985 over 100 will have been studied. To help achieve the NICHD goal, we propose identifying at least 20 subjects per month who are 8-11 weeks pregnant, have viable pregnancies on ultrasound, and have no factors predisposing to fetal wastage (Form A). After randomization into two groups (chorionic villus sampling vs amniocentesis), we will detail medical history, prior pregnancies, genetic history and socioeconomic factors (Forms B,C,D). Subjects randomized in the chorionic villus group will undergo the procedure at 8-11 weeks. Ultrasound will be performed one week thereafter, and again at 16 weeks. Pregnancy complications will be recorded at 4-week intervals throughout gestation. Maternal stress will be assessed upon entry, at 24 weeks and upon termination (delivery or fetal loss). The amniocentesis group will undergo that procedure at 16 weeks, having an ultrasound one week later and surveillance further, as described above. All fetal losses will be subjected to chromosomal studies and anatomic dissection, with DNA analysis or metabolic assays if appropriate. Placenta and cord will be thoroughly examined (Form J) and within two weeks major and minor anomalies determined by geneticists (Form K). Neonatal development will be assessed at nine months, and long-term maternal complications at one and two years. Diagnostic accuracy and laboratory quality will be on the basis of recording technical details (Forms L and M). Finally, women insistent upon CVS will be followed to allow comparison with those agreeing to randomization. Provisions for recruiting controls are made, should attempts at randomization be unsuccessful. All data will be recorded on standard forms, 16 of which are provided in detail in this proposal.